Forget the Dynamic Campaign

At its present status LockOn phisically can't afford a Dynamic Campaign, intended as a dynamic war going on (as it was in Falcon 4).

This evening I started to build a mission, a Day One CAP in a Central European style scenario, then I began to add things.... so at the end I came up with these units:

Most flights had 4 aircraft, some had just 2.

The ground units had about 10-20 vehicles each.

So there were about 100 aircraft and 200 vehicles. (but not all aircraft were scheduled to appear in the first minutes of the mission)

Each unit had its route and its targets defined.

While still in the mission planning phase the FPS began to drop down (less than 70 FPS, instead of more than one hundred, usually, maybe even more)

As soon as the mission was loaded I felt the FPS were close to zero. I could notice really little difference between the paused and the non-paused status of the game.

Even switching to the theatre map view (F10) the situation was still very sad, with at most 9-10 FPS, but usually down to about 5.

The graphics were set about to medium, and my PC config is:

XP-3200+, Nforce2, 1 GB RAM @ Cl2.5, Radeon 9800Pro@XT.

As you can see my pc is not the state of the art, but it's a good machine.

And while I was in the cockpit I had 0 (zero) FPS in this crowded scenario, without any dynamic campaign going on (just units following planned orders). The fact that FPS were so low even in the "map view" tells that it's not just a matter of graphics, it's the game engine which can't stand this number of fighting units. And this without all the "AI umbrella" which is above each dynamic campaign....

Given these performance, I'd say that a dynamic campaign in LockOn today would be impossible to implement, unless the dev rebuild the game engine from scratch.

I would think that placing all these missions as part of a Campaign engine would be more efficient then doint it as you did. I don't know that for fact, but just thinking out loud.

Re: Forget the Dynamic Campaign

Important here to clarify exactly what a "dynamic campaign" means.

Technically, a mission with five units, so long as the results are carried forward and used to build the next mission through a generator, is dynamic.

What you're talking about is the huge immersive theatre, which isn't necessarily the same thing . . . . you can have either of these things without the other.

You are correct, though - Lomac as it is now, and hardware as it is now, can't do the huge numbers of AI units.

ED have stated that they're working on a dynamic campaign for the "next project" . . . . what else they do to ensure that it's reasonably practical will determine whether it's a success or not.

Only time will tell what they can do with future developments :)

I would think that placing all these missions as part of a Campaign engine would be more efficient then doint it as you did. I don't know that for fact, but just thinking out loud.

Maybe... but keep in mind that a dynamic campaign must also "think" at both strategic and tactical level. The AI must decide whether or not to launch a mission, which kind of mission, when and where. And it should also keep track of all data such ammo, spares, fuel and so on.

I remember Falcon 4 was very slow on my early machine (a K6-200 with 64 MB of RAM and a 3DFX card) but it scored about 10-15 FPS while on a dynamic campaign. This time I got FPS ranging from 0 to 2, without the AI thinking at strategic level, just at tactical level. And, another thing... a couple of hundred vehicles are not that much, in a modern campaign. Anyway I'd be happy even if LockOn could manage a dynamic campaign in a small-scale conflict simulation.

Re: Forget the Dynamic Campaign

Important here to clarify exactly what a "dynamic campaign" means.

Technically, a mission with five units, so long as the results are carried forward and used to build the next mission through a generator, is dynamic.

Maybe it's not a complete definition. In a dynamic campaign it would be nice not only to carry forward the results of a single mission, but also to have the AI decide on the battlefield, reacting to events. This is my definition of dynamic.

Your kind of dynamic campaign is already being developed by a team in their own addon for LockOn.

What you're talking about is the huge immersive theatre, which isn't necessarily the same thing

Yes, sure. A dynamic campaign could also be applied to smaller scenarios, but it will lose something, IMHO. A dynamic campaign is cool if you can see at strategic level the effects of what you're doing. Example: If you fail to perform some CAS missions, enemy tanks overrun your airfield; if you cut the enemy supply lines, you halt their advance and so on... IMHO a dynamic campaing should give a deeper insight into the battlefield.

The smaller the scenario, the smaller is the randomization level of the simulation, imho...

Well, to me, the standard is Falcon 4's DC. So if Falcon 4 can do it, with the DC it has which is less than perfect but certainly the standard, and you ran it on such a low end machine, why couldn't LOMAC make the same decisions on the machines we have now which are 6-7 times the machines of yesteryear? Granted it used the bubble method, but I think it is quite doable once hardware catches up to LOMAC. EECH has a really good campaign generator. MIG Alley is really good. All came out several years ago. Maybe if things were optimized on the graphics to maximize FPS, without taking away the beauty, then an engine like that can be done.

I think once hardware starts catching up to LOMAC then a DC of some sort and scale would completely be possible, though I think there would have to be more terrain added to make it worthwhile.

As you know... the Lock On engine renders and tracks all objects at 100% LOD 100% of the time... all sensors, LOS, AI ,flight models, vehicle movement, and explosions...100% of the time.

I believe the bubble system in F4 only tracks those objects within the bubble and the rest is smoke and mirrors... sort of.

Hopefully in future Eagle work we'll see a system that is much more efficient and uses much less resources for the objects that are out of sight.

EECH has a really good campaign generator. MIG Alley is really good.

They are different types of generated campaigns.

The closest likely for the current LOMAC engine is the following:

A map of the front is held on the hard-drive as well as the Squadron Info.

It's loaded, a mission is generated and along the path is arrayed the usual front line activities in a narrow sector.

After the mission is completed, the damage to the enemy and the squadron is calculated and the map of the front is adjusted for:

*Percentage of Effectiveness

*percentage of Supply needed to advance

*Territorial gains by land troops(If in a mission the ground forces were to arrive)

*and the stategic control of various choke-points - Weakness/Strength

For the current LOMAC a non-bubble, off-line mission generator is all we can expect.

The model to examine would be the DCG from IL2. It appears to be the most currently successful DC being done. It has it's faults but it is doable even with all the eye-candy Oleg could put up.

Thats exactly why the FPS falls to such levels if you have too many units involved! Its important for a DC to focus the FPS on areas of importance and make its own assumptions on the rest. At the end of your mission it calculates then the results onto the map. If the pilot strays into areas outside his zone then the units on the map will show up and then those hes already left outside his zone are no longer calculated. If you hit external views only the objects in your zone should show up.

cheers

Subs

yes anyone famiiliar with the F4 dynamic campaign system will know about the infamous bubble. The actual units that gets rendered are only in your bubble, everything else is short circuited until it gets into the bubble. I believe the lomac engine is a bit different in that respect. If you put that many units in F4's bubble it will slow to a crawl too.

I agree with you tho, a dynamic campaign system is one that is written from scratch, unless they go back and redesign it the only dynamic campaign lomac can achieve would be the one in janes f18, not falcon4

A dynamic campaign has nothing to do with the numbers of units in each mission, nor what they do within the mission. That's AI routine.

Falcon 4's campaign is the ultimate one, sure it would be nice to have something like that for Lock On, but it's just one of many many variants in several kinds of simulations.

Here is a definition of a dynamic campaign: "A campaign system where new missions are generated based on set factors and/or results from previous missions."

Full stop.

You can do 70% of the gameplay value in the interface. You don't have to account for single ground units at all. Figuring out every possible thing that can happen in a real war does not help you get a good computer entertainment product.

Some simple points:

- The strategic overview between missions tells the player about the war. Not the group of fighter bombers attacking a column of trucks 250km away from you.

- Distance between events builds tension. A crowded map builds frustration.

- There is no action, event or response that needs to span more than 2 hours. If you stay airborne for more than 2 hours, the map dies out. Realistic? No. Reasonable compromise? Yes.

- Apart from more nuanced AI routines and a simple scramble task, nothing more needs to be added to Lock On's in-mission features to facilitate a dynamic campaign. It does its job between missions, not in them.

- If you are calculating supplies/reinforcements with more than two modifyers, it's too big.

I could go on. It's not about modelling every sock and spoon of two battling armies. It's about maximum gameplay for minimum bloat.

I think once hardware starts catching up to LOMAC..

Lomac should stop running and go back to where the hardware is :lol:

It's just a matter of vastly improving the efficiency of the campaign and AI algorithms. A well designed dynamic campaign engine would fly on today's monster machines.

A dynamic campaign has nothing to do with the numbers of units in each mission, nor what they do within the mission. That's AI routine.

We've already agreed on that... but the point is that to make a mission with 4 aircraft and 4 tanks you don't need a dynamic campaign ;) it's not worth the effort!

Falcon 4's campaign is the ultimate one, sure it would be nice to have something like that for Lock On, but it's just one of many many variants in several kinds of simulations.

Falcon 4's campaign is still the best out there, it was programmed 8 years ago. Also other sims (maybe Combat Pilot for the Amiga) had dynamic campaigns going on during the game. I think that asking less for a modern simulation will make the attribute modern not appropriate.

Here is a definition of a dynamic campaign: "A campaign system where new missions are generated based on set factors and/or results from previous missions."

Full stop.

That's your definition! Your kind of DC is already being explored by the folks @ lockoncampaign. My kind of DC is that where the campaign is dynamic even when I'm flying... I don't wanna have to issue orders to all the units, both allied and enemy, it's up to the AI to acknowledge the objectives and try to achieve them.

Figuring out every possible thing that can happen in a real war does not help you get a good computer entertainment product.

:) :) :) :) :)

Falcon 4 is alive and kicking after 7 years from its release! And its graphics sometimes are far from being photorealistic as in lockon. The reason why some people still prefer Falcon4 (and Falcon4 has a community that is several times that of Lomac)? A dynamic campaign where you feel involved like in a real(simulated) war, and the ability to fly in each and every aircraft that's been around in the last 40 years!

Lockon has great graphics that make you feel you're flying the real thing, but as soon as you have something to do with your wingmen or other AI units you're called back to reality and you see you're flying in a game.

- The strategic overview between missions tells the player about the war. Not the group of fighter bombers attacking a column of trucks 250km away from you.

I don't agree... I LOVE in Falcon 4 when you take off and see other aircraft returning home from their mission, pylons empty and maybe with some battle damage. It gives a sense of immersion which is astonishing.... The number of things that can happen in such a dynamic campaign is incredible... that's why many people enjoy it!

And tell me, what gives most strategic view than other aircraft doing other jobs that interact with yours... single missions put together, if they are not dynamic, are still still single missions, not a DC. Ok you can have the AI remember that a bridge was blown up, but that's all...

- Distance between events builds tension. A crowded map builds frustration.

Do you think that in wartime you can decide where and when to face the enemy? Even in Desert Storm, when Coalition Forces dominated the Iraqi skies, there were times when bomber pilots were threatened by Iraqi fighters. In every conflict THERE IS chaos and frustration. That is why there are civilian casualties, that is why some cheap and old weapon systems often defeat newer and more advanced ones. That is what makes war different from training.

- There is no action, event or response that needs to span more than 2 hours. If you stay airborne for more than 2 hours, the map dies out. Realistic? No. Reasonable compromise? Yes.

Well, a ground attack may last for more than 2 hours, and may require continuos CAS. What would you do after 2 hours? You radio the enemy and tell them "Ok folks, let's have a break! Let's take a cup of tea, it's nearly 5 p.m.!"?

- Apart from more nuanced AI routines and a simple scramble task, nothing more needs to be added to Lock On's in-mission features to facilitate a dynamic campaign. It does its job between missions, not in them.

Well to decide a scramble you must have something that thinks and that thinks strategically. "I don't want them to achieve air superiority, so I have to scramble my interceptors". And when you scramble, which aircraft are you gonna send? Armed with which weapons? (one of the fact that most people ignore is that after a few days/weeks of fighting you often don't have much stocks of modern weapons left)

BTW, if you only require that enemy interceptors scamble and SAMs fire at you, in a simple action-reaction dynamic, you can play F-15 Strike Eagle II from Microprose (1990). It seems to implement your kind of dynamic scenario.... ;)

- If you are calculating supplies/reinforcements with more than two modifyers, it's too big.

I could go on. It's not about modelling every sock and spoon of two battling armies. It's about maximum gameplay for minimum bloat.

But that's the beauty and the might of a real DC! you can cut supplies, you can destroy depots, you can cut routes.... wars are not only fought taking out each and every piece of armor and aircraft! If you have "missions with memory" you'll just fly them as they are now. AI will just remember that a bridge is gone, or that a company of APC has lost 2-3 vehicles. but you would still fly dumb missions....

My idea of DC is very close to that of Falcon 4, and from a sim that is 7-8 years older than that I would expect to see it improved, not scaled down. I would pay $ 100, even $ 200 for a LockOn with a Falcon4 (or better) dynamic campaign. I'm not sure I would spend even $ 20 for a LockOn with just a bunch of "missions with memory".

In 3 to 5 years time we will see LOMAC at its full potential. Right now we are just scratching the surface.

And Loving every minute of it !

(You can quote me on that)

Falcon 4 is alive and kicking after 7 years from its release! And its graphics sometimes are far from being photorealistic as in lockon. The reason why some people still prefer Falcon4 (and Falcon4 has a community that is several times that of Lomac)? A dynamic campaign where you feel involved like in a real(simulated) war, and the ability to fly in each and every aircraft that's been around in the last 40 years!

 

Lockon has great graphics that make you feel you're flying the real thing, but as soon as you have something to do with your wingmen or other AI units you're called back to reality and you see you're flying in a game.

You forgot to mention one 'little' detail: F4 had a leaked source code and basically community was able to do all sorts of things with it.

Without this, I doubt F4 would last like, 7 years.

You forgot to mention one 'little' detail: F4 had a leaked source code and basically community was able to do all sorts of things with it.

Without this, I doubt F4 would last like, 7 years.

I don't want to make yet another thread Lomac Vs F4. One of the post said that "Figuring out every possible thing that can happen in a real war does not help you get a good computer entertainment product." IMHO it's wrong, because F4 has gone in the opposite direction and now has a large community, 7 years after its publication. I don't remind any other sim played seriously all this time. The factor that you can add things as you like is important but imho the fact that it's not limited to single missions is important too.

And remember that no one has forbidden ED to add a dynamic campaign to LockOn or to distribute an SDK to create small add-ons, or to make a sim editable like Strike Fighters or FS2004.

Lomac is more sim than game. In fact it has very little game and IS the best flight simulator out there right now. What I really want is an improved mission editor and not a dynamic campaign

Really for a DC you want a multirole aircraft. The major problem with Lockon at the moment is FPS when there is too many objects in sight of the player and too much happening at once. This contributes to very slow FPS, the other problem is ai for air units in particular helos and wingmen. It would be nice to have DC but I couldn't see one made unless these faults got sorted out first.

That's your definition! Your kind of DC is already being explored by the folks @ lockoncampaign. My kind of DC is that where the campaign is dynamic even when I'm flying... I don't wanna have to issue orders to all the units, both allied and enemy, it's up to the AI to acknowledge the objectives and try to achieve them.

 

:) :) :) :) :)

Falcon 4 is alive and kicking after 7 years from its release! And its graphics sometimes are far from being photorealistic as in lockon. Lockon has great graphics that make you feel you're flying the real thing, but as soon as you have something to do with your wingmen or other AI units you're called back to reality and you see you're flying in a game.

But the creators of F4 are dead. Tell us about the resources and money required to create your Dream Full Theatre DC. How long did it take and how many millions of dollars to get a version of F4 to a trust-worthy state?

The LOMAC engine, under the current marketing arrangements, can't tap a Western Market to pay for the work for a Full Theatre DC.

I would prefer ED to lock up a good missions-based dynamic campaign(Like DCG) before possibly bankrupting themselves for your idle pleasure.

So the Priorities for the current LOMAC engine should be:

• Robust Wingman AI - including Element AI
  Stronger Ground AI
  a good missions-based dynamic campaign

Fulfilling the above would give us a long-term sim worthy of being a long-term gem like LongBow2 which had a limited DC campaign.

Well, to me, the standard is Falcon 4's DC. So if Falcon 4 can do it, with the DC it has which is less than perfect but certainly the standard, and you ran it on such a low end machine, why couldn't LOMAC make the same decisions on the machines we have now which are 6-7 times the machines of yesteryear?

This pretty much sums it up to me.

That falcon4 can do it shows that using the right methodology it is possible.

The changes needed on the AI control side of things isn't too difficult - using the same bubble method as falcon - two layers of AI - call them tactical and strategic. Even without creation of a dynamic campaign implementing such a system of AI control would very much improve the performance of the sim. I think the F4UT site or maybe some falcon forums have a detailed explanation of bubbles and how they work. Therefore the concept is already made, a practical implementation has been demonstrated.

I think a lot more effort would be required to make a working dynamic campaign engine as you are essentially making an old fashioned wargame.

But the creators of F4 are dead. Tell us about the resources and money required to create your Dream Full Theatre DC. How long did it take and how many millions of dollars to get a version of F4 to a trust-worthy state?

How much of the F4 development effort was spent on the campaign engine? Do you know?

How much did F4 cost to develop? Do you know?

How rude and unpleasant you are.

But the creators of F4 are dead. Tell us about the resources and money required to create your Dream Full Theatre DC. How long did it take and how many millions of dollars to get a version of F4 to a trust-worthy state?

 

How much of the F4 development effort was spent on the campaign engine? Do you know?

 

How much did F4 cost to develop? Do you know?

 

How rude and unpleasant you are.

The F4 dynamic campaign system was horribly broken when it came out. Even the developers last patch (1.08) didn't fix it. The wall of red, airbase relocation bugs, missiles fired from outside the bubble couldn't be detected, resource problems, same mission generated over and over again, suicide missions and it goes on and on. Microprose went bankrupt, the developers were fired, someone leaked the source which was like a huge help because exe edits and source edits became possible, the community really pulled it together over the last 3 years to fix all the bugs.

A lot of effort by the community, time, luck and some illegal source leaks is what makes the F4 DC what is now. Many games aren't that luckly. Parent company gets tired of all the whining, releases the developers and locks the code in the vault forever.

Bah I was about to write an intelligent retort, pointing out how much harder it is for a community-based effort to fix a broken sim than the original developers, when I realized that hey - nothing I say can make a difference to those who have already made up their minds, I'm up against religion here.

I'll let my parting words be these:

Bah!

But the creators of F4 are dead. Tell us about the resources and money required to create your Dream Full Theatre DC. How long did it take and how many millions of dollars to get a version of F4 to a trust-worthy state?

 

How much of the F4 development effort was spent on the campaign engine? Do you know?

 

How much did F4 cost to develop? Do you know?

 

How rude and unpleasant you are.

I know it cost 4 years and millions of dollars estimated at least 2 million. I've heard other higher numbers. It was the last big-budget flight-sim that took 4 years to complete.

I know a great portion of the development was spent on the AI - that includes the individual object AI(radar, SAMs, AWACS and the dynamic portion of F4). And from all the patching that was done by the community afterwards, it was not done well. And that community patching - hundreds of free hours - how many $$$ is that worth to get a truly polished product?

But how does knowing these things change the fact that there are more resources(money and time) needed for the full theatre DC.

A single programmer can create a simpler mission-based DC as was done for IL2(with DCG). But the full theatre DC would require more people due to having to test all the various combinations that come up. At both Tactical(A2A AI and Wingman Responses) and Strategic levels(supplies, Troop quality and planning) and tied together with a functional interactive Map.

When it comes to Age of Empires and the like, how much time is spent in the testing and polishing stage? With a full theatre DC, you are adding an RTS to a flight sim with plane(s!) that has to model all its sub-systems correctly.

While it might be difficult for you to get past my tough 1st few sentences, I did present a reasonable destination that would keep LOMAC afloat and might give us a pathway to a reasonable DC product - maybe even a full theatre in the long term(the project after LOMAC).

So the Priorities for the current LOMAC engine should be:

• Robust Wingman AI - including Element AI
  Stronger Ground AI
  a good missions-based dynamic campaign

Stronger AI - both air and ground - are good building blocks for your desired full theater DC. And even a DC at the level of the DCG would keep a community(off-line) interested.

DC = new sim

Dynamic Campaign like in F4 should've been in there from the start.

Like in EECH/EEAH the game is built around it... or not.

If that is ED's aim (don't think so) they might as well start all over again AFAIK.

Indeed F4 is an effort that might turn out quite unique in the flight sim universe (it already is until now).

Let's be satisfied with the direction ED is heading.

This is a flight sim. At this time not an all out war sim.

I can live with that.

I just love the sorties. The flight itself.

A DC would be nice, but nice scripted missions are just as good.

I am looking forward to FC and the new plane (and damage model).

I am gratefull LO is still alive.

If ED ever decides on a DC, it most probably will be for a new sim.

I can wait for that, no prob (been waitin' years for the likes of LO/FC).